Jun | ZenO

Building @ZenO4AI, Data Layer for Physical AI

[Base @base]

10 reasons to build on Base:

1. Deep liquidity
2. Base Ecosystem Fund
3. Sub-cent fees
4. #1 chain for AI agents
5. Base Batches
6. Trillions in stablecoin volume
7. Largest L2 by TVL
8. 20+ local stablecoins
9. Sub-second blocks
10. Cracked builders

[Jun | ZenO @junh_zeno]

@base @ZenO4AI

[David Tso (dave.base.eth) @davidtsocy]

The next generation of global finance is open, transparent, and built on @base

[Jun | ZenO @junh_zeno]

@ZUUAtqCvqoKX1vR @ZenO4AI Please open a ticket on Discord for inquiries related to the Data Expert role.

[ZenO @ZenO4AI]

https://t.co/0rI8uKm0aS

[ZenO @ZenO4AI]

The foundation is taking shape

We’ve been building the core of ZenO how human actions turn into meaningful signals, and how contributors stay in the loop.

That foundation is nearly ready.
ZenO Beta is coming soon.🕐

[ZenO @ZenO4AI]

From VLMs to PhysBrain 👀

VLMs (Vision-Language Models) help AI understand what it sees. But most training data is third-person, which doesn’t match how robots see the world.

Robots must act in the physical world so they need first person ego centric data.

That is why large scale datasets including over 10,000 hours of first person footage from real factory workers have already been used for robot training.

Ego centric video significantly improves a robot’s ability to perceive reason and plan tasks.

First person data enriched with task specific questions answers and reasoning is essential for Physical AI.

ZenO sits at the center of this shift enabling the transition from VLMs to PhysBrain.

[ZenO @ZenO4AI]

ZenO MVP launches in December 2025.

The first version of ZenO will introduce a new way to capture, contribute, and work with first-person data for AI.

We’re getting closer. ⏳

[Jun | ZenO @junh_zeno]

@mushy OMG 😚

[Jun | ZenO @junh_zeno]

@0xMiden GM🫡

[Jun | ZenO @junh_zeno]

Do you know ____? 👀

The OECD report from 2024 highlights that data is the core driver of AI competition.

AI performance depends on the scale diversity and quality of data and large companies control massive datasets while startups cannot access them. The OECD calls this the data network effect.

High quality data is essential for AI safety and fairness and especially critical for robotics AI. Data scarcity is already limiting growth and data is becoming a scarce strategic resource.

ZenO offers a real alternative to the problems of today’s data markets, from regional discrimination against workers to unethical data supply chains.

Keep watching what we build next 🫡

[Jun | ZenO @junh_zeno]

7/ ZenO is not just a data platform.

It is a global network for a new era of intelligence.
Anyone in the world should be able to contribute data without contracts or friction. Traditional data collection simply cannot scale to this level.

ZenO collects real world object understanding, interaction patterns, affordances, and task logs.
For robotics AI companies, it is hard to imagine anything more valuable

[Jun | ZenO @junh_zeno]

Let us go a bit deeper today 👀

ZenO is a data collection platform for robotics AI.
So what kind of data do robots actually need?

[Jun | ZenO @junh_zeno]

6/ We capture this across objects, environments, and human actions and deliver it in structured form to companies.

This is a domain that cannot be solved by foundation models like GPT. And this is exactly why ZenO is necessary.

[Jun | ZenO @junh_zeno]

OpenAI and Meta have been outsourcing AI data work to developing countries.

They paid outsourcing companies about $12 per hour, but the actual data workers received less than $2.

These workers had to work under unstable contracts with very low pay.

ZenO follows the spirit of blockchain by removing middlemen and working directly with the workers. We provide fair rewards based only on performance, not on location or region.

It is time to think seriously about ethical AI data supply chains. Data annotators around the world will stand with ZenO..🫡